Various methods of detecting particles in air are known. One method involves projecting a beam across a monitored area and measuring the attenuation of the beam. Such detectors are commonly known as ‘obscuration detectors’, or simply ‘beam detectors’.
Some beam detectors employ a co-located transmitter and receiver with a distant reflector, and others use a separate transmitter unit and receive a unit located on opposite sides of the open space being monitored.
An exemplary, conventional beam detector is shown in FIG. 1. The detector 10 includes a light source and detector 12 and a reflector 14 placed either side of a monitored area 16. Incident light 18 from the light source and detector 12 are projected toward the reflector 14. The reflector 14 reflects the incident light 18 as reflected light 20 back toward the light source and detector 12. If particulate matter enters the monitored area 16, it will attenuate the incident light 18 and reflected light 20 and cause the amount of light received at the light source and detector 12 to diminish. An alternative beam detector separates the light source from the detector and omits the reflector and directly illuminates the detector with the light source across the monitored area 16. Other geometries are also possible.
Whilst the mechanism of smoke detection used by beam detectors is sound, beam detectors commonly suffer from a number of problems.
Firstly, beam detectors may suffer a type I (false positive) error where foreign objects or other particulate matter, such as dust, enters the monitored area and obscure the beam. Beam detectors are generally unable to distinguish between the obscuration caused by particles of interest e.g. smoke, and obscuration which results from the presence of foreign body of no interest e.g. a bug flying into the beam.
Secondly, beam detectors may require careful alignment at the time of installation. Such alignment aims to ensure that in normal conditions, free from particles, light enters the sensor so as to capture the majority of the transmitted beam, and to in turn maximise sensitivity to an obscuration. This calibration may be slow and therefore costly to perform. Moreover, it may need to be repeated as the physical environment changes, for example because of small movements in the structure to which a beam detector is attached. In some cases, if the intensity of incident light on the detector diminishes quickly this misalignment may also cause a false alarm.
The inventors have proposed a system to address some of these drawbacks in Australian provisional patent application 2008902909, filed 10 Jun. 2008 in the name of Xtralis Technologies Ltd and International Patent application PCT/AU 2009/000727. An exemplary embodiment described therein and reproduced as FIG. 2 herein includes a light source 32, a receiver 34, and a target 36, acting in co-operation to detect particles in a monitored area 38. The target 36, e.g. a corner cube reflects incident light 40, resulting in reflected light 32 being returned to receiver 34. In the preferred embodiment the receiver 34 is preferably a video camera or other receiver having an array of light sensors e.g. one or more CCD (charge-coupled device) image sensors, or CMOS (complementary metal-oxide-semiconductor) image sensors, or indeed any device capable of recording and reporting light intensity at a plurality of points across its field of view.
In this system the receiver 34 receives all of the light in its field of view 40, and includes imaging optics to form an image of a field of its view 40, including the target 36 on its image sensor. Receiver 34 records the intensity of light in its field of view, in the form of data representing the image intensity at a series of locations throughout the field of view. A portion of this data will correspond, at least partially, to reflected light 42. A microcontroller 54 analyses the image data, and determines which portion of the data provides the best estimate of reflected light 42. Because the receiver 34 has a wide field of view and has the ability to independently measure light at a wide range of points within this field of view the light source 32 need not be carefully aligned with target 36, or with receiver 34, since the effect of a misalignment will simply be that a different portion of data, corresponding to different pixels within the view, will be used to measure the reflected light 42. Accordingly, provided that the field of view of the receiver includes target 36, one or more regions of interest within the image will include a measured value for the reflected light 42.
If smoke or other particulate matter enters monitored area 38, it will obscure or scatter incident light 40 or reflected light 42. This obscuration or scattering will be detected as a drop in the intensity for received reflected light 42 measured in the image region determined by the microcontroller.
Pixels falling outside the region selected by the microcontroller, to include the reflected light 42, can be ignored as light received by these pixels does not correspond to the reflected light 42.
Over time, as the building moves or other factors alter the geometry of the system, the target 36 will still be in the field of view of the receiver 34 however, the image of the target 36 will appear at a different point on the image detector of the receiver 34. In order to address this motion, the microcontroller can be adapted to track the image of the target 36 across its light sensor over time to enable a smoke detection to be performed on the correct image regions over time.
In some embodiments described therein the target 36 is illuminated at two (or more) wavelengths λ1 and λ2 e.g. an infrared (IR) and ultraviolet (UV) wavelength which are emitted by corresponding light sources (or a common source) along two substantially collinear paths.
The wavelengths are chosen such that they display different behaviour in the presence of particles to be detected, e.g. smoke particles. In this way the relative change in the received light at the two (or more) wavelengths can be used to give an indication of what has caused attenuation of the beam.
Furthermore, the applicants earlier application depicts an embodiment capable of monitoring multiple targets simultaneously. According to this embodiment, illustrated in FIG. 3 herein, the detector 50 includes a light source 52, a receiver 54, a first target 56, and a second target 57 acting in co-operation to detect smoke in monitored area 58. Target 56 reflects incident light 62, resulting in reflected light 64 returning to receiver 54. Target 57 reflects incident light 65, resulting in reflected light 67 returning to receiver 54. As with the previous embodiment, the receiver 54 communicates the image data to a microcontroller 74. Microcontroller 74 analyses the data, and determines which portion of the data contains information most strongly related to reflected light 64 and reflected light 67 respectively. At the conclusion of this decision process, the microcontroller 74 will have selected two portions of data, corresponding to respective individual pixels or respective groups of pixels read from its image sensor, that can most reliably be used to measure the intensity of reflected light 64 and reflected light 67 respectively. In this way the system 50 can, by the addition of only an additional target or light source, perform the function of two beam detectors.
Using such a system, present inventors have previously proposed a particle detection system which addresses the seemingly contradictory requirements of the need for high sensitivity and the need for a wide angular range of operation in a beam detection system. However, these constraints as well as constraints on the intensity of light sources able to be used as transmitters mean that there may be a need to further enhance the particle detection system in these respects.
In beam detectors the transmitted light intensity may be limited. For example, there may be budgetary considerations which mean that a relatively low power light emitter must be selected in the product. Furthermore, in some cases, a limited electrical power supply is available, especially if the transmitter unit is powered by a battery. Eye safety is also a factor in limiting the transmission power of the light source as is the potential nuisance effect of visible light from the transmitter. For any of these reasons, a relatively low transmitted signal power may be used in a beam detector. Consequently, the signal to noise ratio of the system may be compromised.
In order to operate satisfactorily whilst keeping the emitted power as low as possible it is advantageous, for sensitivity purposes, that the polar emission pattern of the transmitter and the viewing angle of the receiver are kept as narrow as possible. However, for installation and alignment purposes it is advantageous that the same angles are kept as broad as possible. Accordingly, accommodating these seemingly contradictory requirements of the system can present problems.
A further problem that may arise in such a system is that a reflective surface may provide one or more unintended light paths between the transmitter and the receiver, and so interfere with either the recognition of the direct light path, or cause uncontrolled and unintended contributions to the received signal(s), or both. This effect is exacerbated if the reflective surface is subject to any changes, such as movement with temperature or building wind loads; or the movement of people or vehicles that causes its reflected contribution to vary over time.
Since beam detector components are often mounted just below a substantially flat ceiling, this type of undesired reflection may be common. It has been realised by the inventors that to cause such an issue, the finish of the reflective surface does not need to be obviously reflective or mirror-like, and that even a common matt-painted surface may provide a relatively strong specular reflection at the narrow angle of incidence, such as would typically occur in a beam detector with a long span mounted near a surface. While a mirror like, or gloss finish is the extreme case, even an apparently rough surface may give enough specular reflection to create these problems.
Adjacent walls, particularly glazed walls, may also create a similar issue with the additional complication that blinds or open-able windows may be used at various times. However, this issue does not arise as commonly, since it is rarely required that beams are directed in close proximity to walls.
For this reason and others, beam detectors typically require careful alignment at the time of installation. Such alignment aims to ensure that in normal conditions, free from particles, light enters the sensor so as to capture the majority of the transmitted beam, and to in turn maximise sensitivity to an obscuration. This calibration may be slow and therefore costly to perform. Moreover, it may need to be repeated as the physical environment changes, for example because of small movements in the structure to which a beam detector is attached. In some cases, if the intensity of incident light on the detector diminishes quickly this misalignment may also cause a false alarm.
Since beam detectors are typically mounted to a wall or like flat surface it is generally not possible to get behind the detector in order to use a line of sight type alignment device. Also, since detectors are usually mounted at high elevations and in inaccessible locations, the problem of achieving accurate alignment, and the difficulties caused by misalignment, are exacerbated.
As discussed in relation to FIG. 1, some beam detectors employ a co-located transmitter and receiver with a remote reflector. Another arrangement, as illustrated in FIG. 9, uses a light source 1102 that is remote from the receiver 1104. The separate transmitter 1102 may be battery powered in order to avoid the requirement for costly wiring. Furthermore, in embodiments that are powered from the fire alarm loop the detector unit 1104 (or the combined light source and detector 102, of FIG. 1) may also employ a battery to act as a reserve supply for periods of high power consumption that exceed a specified limit of capacity of a wired loop supply.
In order to achieve the required service life, and for conformance with safety requirements, it is desirable that the battery-powered units should not be powered on during shipping or in long-term storage.
Conventionally, battery-powered equipment is often activated using a manual switch, or by removal of an insulating separator, or by inserting the batteries into the equipment. The inventors have identified that these methods have several disadvantages, particularly in the case of beam-detection systems. The conventional systems for powering up the battery-powered equipment are not automatic and, in consequence, may be overlooked when the beam-detection system is installed. In beam-detection systems the wavelengths used for the light source 102, 202 are often invisible to the human eye. This makes it difficult to confirm that the light source 102, 202 is active when installed. In addition, the beam detection systems are often installed at a significant height, requiring scaffolding or a cherry-picker to access the system components. As a result, it is time-consuming and costly to access and rectify a unit that has inadvertently been left non-operational.
Some of the conventional techniques of activating battery-powered units also interfere with the common requirement that beam-detection systems should avoid arrangements that cause penetrations through the main enclosure of the unit. It is often the case that transmitters are designed to be resistant to the entry of dust and moisture, and the use of manually-operated switches may makes this isolation more difficult and costly to achieve.
A further problem that may arise with beam detectors is that their exposed optical surfaces may become contaminated with dirt over time. This can gradually reduce the received signal with the potential to be raise a false alarm. Methods to avoid and remove dirt build up on optical surfaces are known, and employed particularly commonly in the field of closed circuit TV security surveillance applications, such as the use of contamination-resistant coatings on viewing windows, protective shrouds, wash-wipe mechanisms and the like.
Also, as described in PCT/AU2008/001697 in the name of Xtralis Technologies Limited, there are other mechanical means for cleaning or avoiding dirt build up on optical surfaces, including methods using filtered clean air as a barrier, or electrostatic guard areas to prevent window contamination. Such methods may advantageously be used for beam detectors separately or in combination with other aspects of the current invention, and each constitute an aspect of the present invention.
With the dual wavelength system described in connection with FIGS. 2 and 3 a variation in the absolute intensities of received light is tolerated to an extent, because a differential measure is used to detect particles in the beam, but relative variation between the wavelengths may create faults or, worse still, false alarms; specifically a relative reduction in the received signal from the UV beam compared to the IR beam may be mistaken for smoke. Thus any wavelength selective build up of contaminants on the optical surfaces can be problematic.
It is a problem in the field of video surveillance, and similar fields which have remotely located optical devices (such as cameras), that insects or other foreign bodies may from time to time land on the exposed surfaces of the optical components of the system and partly or totally obscure the field of view of the optical components. Similar problems may also arise in particle detection systems like beam detectors which are exposed to bugs and other foreign bodies. Accordingly, there is a need to protect components of particle detection systems such as a beam detector and thereby to avoid or minimise false alarms caused by such circumstances.
As described above, some embodiments of the present invention may include separate light emitters in the transmitter which are configured to emit light in different wavelength bands. Most preferably the light emitters are LEDs. Over time the output of the LEDs may vary in either absolute or comparative intensity or both. With the dual wavelength system variations in absolute intensity can be tolerated to a certain extent so long as the relative measure of intensity used by the system for detecting particles remains substantially constant. However, relative variations in the output intensities of the two light emitters may create faults or false alarms. This is particularly the case when the output signal from the UV LED reduces compared to the output of the infrared LED.
It is known to use beam detectors to monitor large areas by using beams over say, 150 meters long or, in relatively confined spaces requiring a beam length of eg. only 3 meters. In conventional beam detector systems an identical light source and receiver can be used for these two very different applications, i.e. 150 meter separation or for 3 meter separation. This is made possible by either adjusting the gain on the receiver or turning down the transmitter power according to the separation between the transmitter and the receiver.
However, the applicant's previous applications discussed above, and the example of FIG. 3 show a beam detector which may include more than one transmitter for each receiver. This presents its own particular problems, in that it is possible to have multiple transmitters set at vastly different distances from the receiver. For example, consider a room of the type illustrated in FIG. 57. This room 5700 is generally L-shaped and has a receiver 5702 mounted at the external apex of the L-shape. Three transmitters 5704, 5706, 5708 are positioned around the room 5700. The first transmitter 5704 is located along one arm of the L. A second transmitter 5706 is located in a position 90° from the first receiver 5704 at the end of the other arm of the L. A third transmitter 5708 is mounted across the apex of the L-shape from the receiver 5702. As will be appreciated the distances between the transmitters 5704 5706 and the receiver 5702 are much longer than the distance between the transmitter 5708 and the receiver 5702. As a result, the level of light received from each transmitter will be very different. Moreover transmitter 5708 may be so close to the receiver that it saturates its light receiving element.
Other disadvantages may also arise, for example, from time to time, an installer may take advantage of the reliable performance of beam detectors and install a system outside the manufacturer's specifications. For example, although beam detectors are often intended to operate with a substantial separation between the transmitter and receiver an installer may extend this distance to provide a system beyond that recommended by the manufacturer or allowed by regulations. In some cases an installer of the particle detector may not know of the limits of operation of the receiver for the light source provided therewith.
In such circumstances an installed particle detector may operate satisfactorily at initial installation, but sometime following installation, cease to operate correctly. This may occur, for instance where the particle detector or was initially installed close to, but beyond its design limits. Over time, changes may occur to the equipment or environment, which gradually alter the received signal strength due to reasons other than the presence of particles in the beam. These changes may be caused by, for example, component ageing, gross alignment drift, or contamination of optical surfaces. Such system drift would ordinarily be handled by the system if it had been set-up within its design limits. However, when the system is set up outside these limits, degradation of performance and the associated occurrence of fault conditions may occur prematurely or repeatedly.
Furthermore, it is desirable to be able to calibrate and/or test such a beam detector by simulating the presence of smoke using a solid object. Such a test is a requirement of standards bodies testing for beam detectors. For example, the European EN 54-12 standard for ‘Biodetection and fire alarm systems. Smoke detectors. Line detectors using an optical light beam’.
In prior art testing methods the testing of beam detectors employs a light filter that partially obscures the projected light beam to simulate the effect of smoke. The filters used usually consist of a mesh of fibres, or dye-loaded plates or transparencies with printed features which obstruct all visible and near visible wavelengths by substantially the same amount in a repeatable fashion. The present inventors have realised that this type of filter may not be suitable for use with a beam detector of the type described above.
In a preferred embodiment of the system described in FIGS. 1 to 3, the light sources are configured to include a plurality of light emitters, wherein each light emitter is adapted to generate light in a particular wavelength band. Moreover, the separate light sources are arranged to emit light at different times in order that a monochromatic imaging element may be used. The direct result of the use of separate light emitters is that there is some separation between the two light emitters in the light source, and thus the light will travel over slightly different, although closely adjacent, beam paths through the intervening space between the light source and receiver. This provides a risk that a small object such as an insect on the transmitter could affect one light path more than the other and so affect the reading of the receiver. This may induce a false alarm or unnecessary fault condition.
Conventional beam detectors require careful alignment at the time of installation. Such alignment aims to ensure that in normal conditions, free from particles, light enters the sensor so as to capture the majority of the transmitted beam, and to in turn maximise sensitivity to an obscuration. This calibration may be slow and therefore costly to perform. Moreover, it may need to be repeated as the physical environment changes, for example because of small movements in the structure to which a beam detector is attached. As stated above, the inventors have previously proposed a particle detector in PCT/AU 2008/001697, filed 10 Jun. 2009 in the name of Xtralis Technologies Ltd (the specification of which is incorporated herein, by reference, in its entirety) which includes a receiver which has a light sensor comprising matrix of light sensor elements, e.g. CCD (charge-coupled device) image sensor chip, or CMOS (complementary metal-oxide-semiconductor) image sensors such as in a video camera, or other receiver that is capable of receiving and reporting light intensity at a plurality of points across its field of view. Each sensor element in the receiver produces a signal that is related to the intensity of the light that it receives. The signals are transmitted to the controller, where a particle detection algorithm is applied to the received image data. Compared to a single-sensor receiver, the receiver in this particle detector has a wider field of view but lower noise and has the ability to independently measure light at a wider range of points within this field of view.
Because each sensor element has an inherent noise level, the overall signal-to-noise ratio of the system can be improved by focusing the target (i.e. beam image) on a single sensor element. However, this may not yield optimal results.
The above mentioned type of sensor e.g. CCD's and the like, are sometimes subject to a phenomenon created by the image processing algorithm used for the receiver, known as staircasing, wherein adjacent pixels or adjacent groups of pixels have significantly different values. The physical structure of the sensor also has non-responsive “gaps” between sensor elements that produce no signal. Because of these effects, any variation in the alignment of the smoke detector components can potentially create a large variation in the measured light intensity level.
For example, because of the small size of the focused target, a very small movement of the receiver or the transmitter could cause the target to move onto an entirely different sensor element with a very different inherent noise level or response compared to the previous pixel on which it was focused. It may also fall into a position, where all, or a non-trivial part, of the received beam falls into one of the aforementioned “gaps”. The resulting variation in the image intensity as determined by the controller can thus potentially cause the controller to falsely detect smoke.
To partly ameliorate this problem, the detector can be adapted to track the target across the light sensors over time to enable a smoke detection to be performed on the signals from the correct sensors over time. However, to properly determine the image intensity, the controller will be required to ascertain the inherent properties of different light sensors used over time. Doing so requires system resources such as processing cycles and power. Also it is not always possible for the controller to make this determination.
In beam detectors an additional problem that may arise is interference from ambient light within the volume being monitored. The ambient light can either be from sunlight illuminating the volume or artificial lighting used to illuminate the space. Accordingly, beam detectors require mechanisms for minimising the impact of this light. This problem is compounded by the conflicting requirement that the light sources of the beam detector should be relatively low powered so that they minimise power consumption, are eye safe and do not create a visible nuisance. In prior art beam detectors which use a single wavelength of light a filter is typically used to reduce the signal from ambient light. In the case of an infrared beam detector this is generally a low pass filter that removes substantially all visible and UV light. However, this is inappropriate for a multiple wavelength system as described herein.
In the preferred embodiment of the system described above the particle detector is powered at the receiver directly from the fire alarm loop. This minimises the installation costs of the device in that it obviates the need for dedicated wiring for supplying power or communicating with the detector. However, the fire alarm loop usually only provides a very small amount of DC electrical power for the detector. For example, an average power consumption of about 50 mW may be desirable for such a detector. However with current technology the power consumed during video capture and processing may be far above the 50 mW that is available from the loop. To address this problem a separate power supply could be used, but this is costly since standards for fire safety equipment are onerous, e.g. they require a fully approved and supervised battery backed supply, and fixed mains wiring.
The limited supply of power also limits the optical power output of the transmitter. The limited optical power output in turn limits the signal to noise ratio of the measured signal. If the signal to noise ratio of the system degrades too far, the system may experience frequent or continual false alarms.
In some systems, the signal to noise ratio can be enhanced by employing long integration or averaging times at the receiver. However system response times, which are usually between 10 and 60 seconds, must be increased to higher levels if long integration times are used. This is undesirable.
In addition to using a beam detector for smoke detection it is often desirable to use other sensor mechanisms for detecting additional or alternative environmental conditions or hazards, for example CO2 gas detection or temperature detection. The detectors conventionally use a wired or radio communication link to signal an alarm or fault condition to fire alarm control panel or like monitoring system. As such these links often add significant cost and potential reliability issues to the alarm system.
In some systems the present inventors have determined that it can be beneficial to operate at least some components, and most advantageously the transmitter on a battery. An exemplary component is described in the applicant's co-pending patent application no. PCT/AU 2009/000727, filed on 26 Jun. 2008, the contents of which are incorporate herein by reference for all purposes.
However, a problem that can arise in a battery powered component of a particle detector is that over time, the batteries of the component will become discharged and the component will ultimately fail. Such failure will potentially require an unscheduled maintenance call out for the device to be repaired and recommissioned. In a smoke detection application this is particularly problematic as the equipment is used in a life-safety role and faults are required to be rapidly remedied. The problem can be remedied by performing preventative maintenance but ultimately this may amount to performing unnecessary servicing and replacement of units that have a significant amount of battery life remaining and therefore is costly and wasteful of materials.
Unfortunately, variations in individual battery performance and environmental conditions make simply scheduling routine replacement periods unreliable and potentially wasteful. One apparent solution to the problem is to equip the component with an indicator of battery state, however this has a disadvantage of adding cost, and the indicator itself is power consuming which further reduces battery life. Moreover, it requires regular direct inspection of the indicator on the component which, in the case of a beam detector, may be particularly inconvenient.
In beam detectors such as that described in relation to FIG. 3 i.e. where a plurality of beam detectors are formed by corresponding transmitter and receiver pairs, such that two or more beams either intersect or pass through a common region of air, sufficiently close to each other that their points of intersection can be mapped to addresses within the region being monitored, a problem may arise in that any one of the subsystems may be affected by environmental conditions or system problems that do not affect the other subsystem. Such issues generally force a reduction in achievable sensitivity or increase the rate of unwanted false alarms.
Reference to any prior art in the specification is not, and should not be taken as, an acknowledgment or any form of suggestion that this prior art forms part of the common general knowledge in Australia or any other jurisdiction or that this prior art could reasonably be expected to be ascertained, understood and regarded as relevant by a person skilled in the art.